An impressive corpus of health sciences library statistics has been accumulated in the last decade relating to variables such as size of holdings, budget, personnel, circulation, reference services, cataloging practices, interlibrary loan transactions etc. Unfortunately, the types of data traditionally accumulated by libraries and captured by survey do not provide what is required for effective decision making. The proposed project will focus upon four basic factors: who uses health sciences libraries, types of library usage, intended areas of impact, and the associated costs. Maximum use will be made of recent advances in library economics and relevant operations research and econometric techniques. The study will involve the gathering of data, for subsequent analysis and modelling, at five clusters of health sciences libraries, representing a total of 80 to 100 libraries. The anticipated outcome will be the provision, for the first time, of quantitative estimates of the numbers and types of users of the nation's health sciences libraries. It will provide quantitative data on the interface between these libraries and the clinical, research and educational functions involved in health care delivery. In particular, it will show how the library channels information to health care practitioners and the dollar costs of maintaining the library-user interface.